Optimising lung cancer screening for individuals from socio-economically deprived communities.

Abstract

Lung cancer remains the most prevalent and fatal cancer globally, often presenting at an advanced stage. Socioeconomic deprivation is strongly linked to lung cancer incidence and outcomes, with deprived populations experiencing higher rates of late-stage diagnosis and lower survival. Low-dose computed tomography (LDCT) screening has the potential to meaningfully reduce this disease burden by detecting it at earlier stages, when it is more amenable to curative treatment. Risk-based targeted screening has the potential to optimise the balance between the benefits and harms of screening, although there is little empirical evidence demonstrating the actual performance of risk prediction models. This thesis evaluates outcomes and risk model performance from screening initiatives in Greater Manchester. This region has one of the UK’s highest lung cancer burdens and significant socioeconomic inequality. To tackle this challenge, risk-based screening was introduced to areas of Manchester with high levels of deprivation. The overarching aim of this work was to inform the optimisation of targeted screening for effective and equitable implementation at scale. I evaluated participation following different invitation strategies in a high-risk population, identifying groups with lower engagement that may be targeted to reduce inequalities. I explored the role and limitations of using primary care smoking status records for targeted invitation, and identify the most efficient criteria to use. Linkage of real-world data sources spanning a substantial follow-up period enabled investigation of how screening performed in practice. I analysed risk prediction model performance in determining screening eligibility, and investigated the potential for risk-based personalised screening intervals. I found that risk models statistically underestimated lung cancer risk but, overall, effectively stratified participants into high- and low-risk groups. Finally, I found that targeted screening had high yield at population level, and likely reduced late-stage disease incidence by one fifth. The methodology used here was novel and could be applied at scale in the UK setting as screening is implemented nationally. These findings provide insights for national screening programmes aiming to balance effectiveness, efficiency, and equity, ultimately contributing to the goals of improving early cancer detection and reducing health disparities.

Date of Award	26 Aug 2025
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Philip Crosbie (Main Supervisor) & Matthew Sperrin (Co Supervisor)